Discovery of pyrano[2,3-d]pyrimidine-2,4-dione derivatives as novel PARP-1 inhibitors: design, synthesis and antitumor activity

Poly(ADP-ribose) polymerases-1 (PARP-1) are involved in DNA repair damage and so PARP-1 inhibitors have been used as potentiators in combination with DNA damaging cytotoxic agents to compromise the cancer cell DNA repair mechanism, resulting in genomic dysfunction and cell death. In this study, we report the synthesis of a novel series of pyrano[2,3-d]pyrimidine-2,4-dione analogues as potential inhibitors against PARP-1. All the newly synthesized compounds were evaluated for their inhibitory activity towards PARP-1 and examined for their anti-proliferative activity against MCF-7 and HCT116 human cancer cell lines. The synthesized compounds showed promising activity where compounds S2 and S7 emerged as the most potent PARP-1 inhibitors with an IC50 value of 4.06 ± 0.18 and 3.61 ± 0.15 nM, respectively compared to that of Olaparib 5.77 nM and high cytotoxicity against MCF-7 with IC50 2.65 ± 0.05 and 1.28 ± 1.12 μM, respectively (Staurosporine 7.258 μM). Compound S8 remarkably showed the highest cell growth inhibition against MCF-7 and HCT116 with an IC50 value of 0.66 ± 0.05 and 2.76 ± 0.06 μM, respectively. Furthermore, molecular docking of the compounds into the PARP-1 active site was performed to explore the probable binding mode. Finally, most of the synthesized compounds were predicted to have good pharmacokinetics properties in a theoretical kinetic study.


Introduction
PARP-1 has received great attention as a promising anti-cancer therapeutic target. 1 PARP-1 is a highly conserved DNA-binding protein and is the most extensively expressed member of the poly(ADP-ribose) polymerases (PARPs) family which is composed of 18 members. They regulate a number of cellular processes including surveillance of genome integrity, cellular differentiation, regulation of gene transcription, inammation, mitosis, cell cycle progression, initiation of DNA damage response and apoptosis. 2 PARP-1 is a known sensor of DNA damage as it is responsible for DNA base excision repair (BER) and DNA single-strand break (SSB) repair. 3 Damaged DNA activates PARP-1 to cleave its substrate nicotinamide adenine dinucleotide (NAD+) and to catalyze the addition of ADP-ribose units to it and to nuclear target proteins to recruit BER components to facilitate DNA repair process and cell survival. [4][5][6][7] Increased PARP-1 expression is sometimes observed in melanomas, breast cancer, lung cancer, and other neoplastic diseases. 8 It has been disclosed that in BRCA1/2-mutant cancer cells, inhibition of PARP1 is synthetically lethal due to their dependence on PARP-1 activity for DNA (base excision) repair and subsequently survival. 9 So PARP-1 inhibitors have shown success when used as monotherapy for treating genetically DNA repair-defective cancers. PARP-1 inhibitors have not only been used in BRCA1/2 decient cancers but also used in combination therapy with DNA-damaging therapeutics to improve their potencies by blocking DNA-repairing process.

Biological evaluation
The biological evaluation was accomplished through testing both enzyme inhibition activity and anti-proliferative activity. The enzymatic activity of the synthesized compounds was assessed against PARP-1.

2.2.1
In vitro PARP-1 inhibitory assay. The synthesized compounds were evaluated for their PARP-1 inhibitory activity. As shown in Table 1, most of the synthesized compounds displayed excellent inhibitory activities against PARP-1 with IC 50 values ranging from 3.61 nM to 114 nM compared to the reference drug Olaparib. Compounds S2 and S7 showed a higher potency than Olaparib. S4, S5, S8 and S10 showed a slightly less potency then Olaparib. Compounds S1, S6 and S9 showed the least potencies. The obtained result suggesting that the presence of a heterocycle fused with the pyrano[2,3-d] pyrimidine 2,4 dione enhances the potency.
2.2.2 Cell proliferation inhibition of synthesized compounds. All the synthesized compounds were further

Molecular docking study
The crystal structure of the ligand Olaparib complexed with PARP-1 (PDB ID: 5DS3) is shown in Fig. 2. In order to explore the binding mode of the designed pyrano[2,3-d]pyrimidine 2,4 dione derivatives, we obtained and analyzed the docked structures of the synthesized compounds within the catalytic site of PARP-1 and compared it with the that of Olaparib.
As expected, the pyrano[2,3-d] pyrimidine 2,4 dione scaffold occupied the NI-site and interacted with Ser904 through a hydrogen bond with the carbonyl group of the ring, also all the compounds interacted with Gly863 through two characteristic hydrogen bonds with the carbonyl group and NH of the ring. Also all compounds showed p-p stacking interactions with Tyr907 and His862. All these interactions were present in the reference Olaparib ( Fig. 2-5).
As shown in Table 3 the docking results were consistent with the PARP-1 enzyme assay where the potent compounds showed good affinity to the enzyme as illustrated with their docking score. Interestingly, compound S2, which had an IC 50 of 4.06 nM, the pyrimidine ring showed two additional interactions, the amino group did a hydrogen bond with Gly863 and SH group did a p-sulphur interaction with His909 ( Fig. 3) while compound S4, which had an IC 50 of 14.94 nM showed only an additional hydrogen bonding with Tyr907 through the amino group of the pyrimidine ring (Fig. 5) and compound S5, which had an IC 50 of 11.07 nM, showed an additional p-sulphur interaction with Tyr907.
Compound S7, which had an IC 50 of 3.61 nM, showed an additional carbon-hydrogen interaction with Ser864 also its phenyl ring laid in a deep hydrophobic pocket lined with the side chain of Asn868, Ile872 and Leu877 within the binding site (Fig. 4). Compound S8, which had an IC 50 of 15.79 nM, showed an additional hydrogen bond with Arg878 through the nitro group while compound S10, which had an IC 50 of 16.16 nM, showed an additional p-alkyl interaction with Tyr896 through its methyl group (Fig. 5).

Structure-activity relationship (SAR)
Based on the observed pharmacological and molecular docking data we can conclude that presence of pyrano[2,3-d]pyrimidine 2,4 dione scaffold is important for interactions with the amino acids present in the NI site of the enzyme, addition of a fused heterocycle resulted in extra interactions with the enzyme and greatly enhanced the activity, also presence of hydrophobic substituent on the ring was favorable due to interaction with the AD site of the enzyme.

In silico ADMET study
Pharmacokinetics properties of synthesized compounds were predicted using ADMET protocol in Accelrys Discovery Studio 4.1 soware. The results of the ADMET study are presented as ADMET-Plot, which is a 2D plot drawn using calculated PSA_2D and A log P98 properties (Fig. 6).
In BBB plot, most of the compounds except S5 and S7 were fallen outside the 99% ellipse. Hence, these compounds may not be able to penetrate the blood brain barrier; hence the chances of CNS side effects are predicted to be low.
In HIA plot, most of compounds fell inside the 99% ellipse, thus estimated to have good human intestinal absorption except for S3 and S8 which showed poor absorption. Aqueous solubility level of most of the compounds was found to be 3 or 2  which indicates low aqueous solubility. The hepatotoxicity level of all compounds was 1. Hence the compounds are predicted to possess hepatotoxicity. Further experimental studies are required to determine the hepatotoxic dose levels. Most of the compounds are predicted as noninhibitors of CYP2D6. Hence the side effects (i.e. liver dysfunction) are not expected upon administration of these compounds.
The plasma protein-binding model predicts whether a compound is likely to be highly bound to carrier proteins in the blood. There is diversity in the synthesized compounds regarding their ability to bind to plasma proteins.
PSA is a key property that has been linked to drug bioavailability. Thus, passively absorbed molecules with PSA > 140 are thought to have low bioavailability. Most of the synthesized compounds have PSA ranging from 94.77-118.63, thus, they are predicted to present good passive oral absorption except for compounds S3 and S8 which had PSA more than 140. The calculated parameters from the ADMET study are tabulated in Table 4 (ESI †).

Conclusion
In conclusion, a series of novel pyrano[2,3-d] pyrimidine 2,4 dione analogues were designed and synthesized based on the characteristics of the catalytic domain in PARP-1. These compounds were evaluated for their PARP-1 enzyme inhibitory activity and cellular inhibitory against MCF7 and HCT116 human cancer cell lines. Compounds S2 and S7 showed a higher potency than the reference Olaparib indicating that the presence of an extra fused heterocycle ring enhanced the activity. Finally, a molecular docking study was performed to investigate the probable interactions with the PARP-1 enzyme.

Chemistry
Melting points of all the new compounds are listed uncorrected and were measured by a Reichert Thermovar apparatus. Yields registered are of the new compounds. The IR spectra were determined using Perkin-Elmer spectrometer (KBr disc), model 1720 FTIR. 1 H-NMR, and 13 C-NMR spectra were done using a Bruker AC-300 or DPX-300 spectrometers. Chemical shis were reported in d scale (ppm) using TMS as a reference standard and the coupling constants J values are given in Hz. The progress of the reactions was determined using TLC aluminum silica gel plates 60 F245. IR, the analysis ( 1 H-NMR, 13 C-NMR and elemental analyses) were done on the Main Chemical Warfare Laboratories, Chemical Warfare Department, Egypt.

PARP inhibition assay
Assay procedure. PARP-1 enzyme inhibition activity was measured for using a colorimetric 96-well PARP-1 assay kit (catalog no. 80580) (BPS Bioscience), according to the manufacturer's protocol. Briey, the histone mixture was diluted 1 : 5 with 1Â PBS, 50 ml of histone solution was added to each well and incubated at 4 C overnight. The plate was washed three times using 200 ml PBST buffer (1Â PBS containing 0.05% Tween-20) per well. Liquid was removed from the wells by tapping the strip wells on clean paper towels. To each well, 200 ml of blocking buffer was added, followed by 60-90 min incubation at room temperature. Then 25 ml of PARP master mixture (consisting of 2.5 ml 10Â PARP buffer + 2.5 ml 10Â PARP assay mixture + 5 ml activated DNA + 15 ml distilled water) was added to each well. Olaparib was used as a positive control. 5 ml of inhibitor solution of each well labeled as "test inhibitor" was added. For the "positive control" and "blank", 5 ml of the same solution without inhibitor was added. 1Â PARP buffer was prepared by adding 1 part of 10Â PARP buffer to 9 parts H 2 O (v/ v), 20 ml of 1Â PARP buffer was added to the wells designated as "blank". The amount of PARP-1 required for the assay was then calculated. The reaction was initiated by adding 20 ml of diluted PARP1 enzyme to the wells designated "positive control" and "test inhibitor control". The strip wells were incubated at room temperature for 1 hour. The strip wells were then washed three times with 200 ml PBST buffer. Then, 50 ml of 50 times diluted Streptavidin-HRP with blocking buffer was added to each well, and the strips were further incubated at room temperature for 30 min. Aer washing the wells three times with 200 ml PBST buffer, HRP colorimetric substrate was added to each well and the plate was incubated at the room temperature until a blue color is developed in the positive control well. Then reaction was quenched with 100 ml per well of 2 M sulfuric acid, and absorbance at 450 nm was determined. Carrier solvents were assayed as negative controls. All assays were performed in triplicate. To determine the IC 50 value for each inhibitor, the average absorbance of each inhibitor concentration was plotted against the log of the concentration of each respective inhibitor and the IC 50 value for each plot was obtained using computerassisted non-linear regression analyses. Data presented are the results of at least two independent experiments done in triplicate. The results of these studies are presented as mean IC 50 (nM). 24 5.2.2 Cell proliferation inhibition assay. We used the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide] (Biomatik, Wilmington, DE) assay to measure cell growth inhibition, which is based on the conversion of MTT to formazan crystals by mitochondrial dehydrogenases. 25 Only in live cells, mitochondrial enzymes can transform MTT into insoluble formazan. Aer incubated with serially diluted inhibitors for 96 h, cell cultures were incubated with MTT solution (5 mg ml À1 ) for 4 h at 37 C. Then discard the medium and DMSO was added to solubilize the reaction product formazan by shaking for 10 min. Absorbance at 492 nm was measured with a microplate reader (Thermo, MK3). Cell viability was expressed as an IC 50 value.

5.3.
In silico studies 5.3.1 Molecular docking study. Molecular modeling simulation study was performed through docking of the target compounds in the binding site of PARP-1 enzyme using C-Docker protocol in Discovery Studio 4.0 Soware. The X-ray crystal structure of Olaparib in complex with PARP-1 was downloaded from http://www.rscb.org/pdb (PDB ID: 5DS3) in PDB format. Computational docking is an automated computer-based algorithm designed to estimate two main terms. 26 The rst is to determine the suitable position and the orientation of certain test set molecule's pose inside the binding site in comparison to that of the X-ray crystallographic enzymesubstrate complex. The second term is the calculation of the estimated protein ligand interaction energy which is known as docking scoring.